Oscillation probability
The chance that a neutrino born as one flavour is detected as another.
P(να→νβ) = δαβ - 4 Σi>j Re(U*αi Uβi Uαj U*βj) sin²(Δm²ij L/4E) + 2 Σi>j Im(...) sin(Δm²ij L/2E). The imaginary terms change sign for antineutrinos and contain CP violation.
Try the oscillation calculator →
In plain words
Every oscillation experiment ultimately compares counts with one number: the probability that a neutrino produced with flavour α arrives as flavour β. This probability is a sum of waves. Each pair of mass states contributes one wave whose frequency is set by their mass-squared difference and whose height is set by the mixing matrix. Real measurements see the superposition of these waves, smeared by the energy spread of the beam and the resolution of the detector.
In detail
For three flavours in vacuum, the full expression contains a CP-even part, which is the same for neutrinos and antineutrinos, and a CP-odd part, which changes sign:
Three properties follow directly. Probabilities summed over all final flavours equal one (unitarity). For α = β the imaginary part vanishes, so disappearance channels cannot see CP violation. All imaginary parts share a common magnitude, the Jarlskog invariant, so CP violation shows up only in appearance channels and requires all three angles to be non-zero.
Example
For the νμ → νe channel at long-baseline experiments, the leading term is sin²θ23 · sin²2θ13 · sin²Δ31, a few percent at the first maximum. Interference with the solar term adds a piece proportional to sin δCP, which raises the probability for neutrinos and lowers it for antineutrinos when δCP is near −90°, the value currently favoured. Matter modifies these terms further, see matter effects.
Why it matters for neutrino physics
The formula is the bridge between theory and every measured parameter. Experiments are designed around it: they choose L/E to sit near a maximum of the desired term, and they use simplifications such as the two-flavour approximation when one term dominates. Global fits use the complete version, including matter, to extract all parameters at once.
Frequently asked questions
Why do neutrinos and antineutrinos have different probabilities?
Only the imaginary part of the mixing-matrix products changes sign between neutrinos and antineutrinos. If δCP is not 0° or 180°, this term is non-zero and the two probabilities differ.
Can the probability exceed 100 %?
No. Unitarity of the mixing matrix guarantees that the probabilities to arrive as electron, muon or tau neutrino always add up to one, as long as there are no additional sterile states.
Original papers
- TheoryV. Gribov and B. Pontecorvo, Neutrino astronomy and lepton charge, Phys. Lett. B 28, 493 (1969)DOI 10.1016/0370-2693(69)90525-5
- ReviewS. M. Bilenky and B. Pontecorvo, Lepton mixing and neutrino oscillations, Phys. Rept. 41, 225 (1978)DOI 10.1016/0370-1573(78)90095-9
Go deeper
- Two-flavour approximationA simple formula for oscillation when only two neutrino types matter.
Related terms
Mentioned in 22 pages
People 5
Last updated: · Editorial standard
Cite this page
Text
Neutrino Atlas (2026). “Oscillation probability”. Retrieved 22 November 2026, from https://neutrino-atlas.com/terms/oscillation-probability/BibTeX
@misc{neutrinoatlas2026msoscillationprobability,
title = {Oscillation probability},
author = {{Neutrino Atlas}},
year = {2026},
url = {https://neutrino-atlas.com/terms/oscillation-probability/},
note = {Accessed 2026-11-22},
language = {english}
}